[deliverable/linux.git] / net / netfilter / nfnetlink.c

/* Netfilter messages via netlink socket. Allows for user space
 * protocol helpers and general trouble making from userspace.
 *
 * (C) 2001 by Jay Schulist <jschlst@samba.org>,
 * (C) 2002-2005 by Harald Welte <laforge@gnumonks.org>
 * (C) 2005,2007 by Pablo Neira Ayuso <pablo@netfilter.org>
 *
 * Initial netfilter messages via netlink development funded and
 * generally made possible by Network Robots, Inc. (www.networkrobots.com)
 *
 * Further development of this code funded by Astaro AG (http://www.astaro.com)
 *
 * This software may be used and distributed according to the terms
 * of the GNU General Public License, incorporated herein by reference.
 */

#include <linux/module.h>
#include <linux/types.h>
#include <linux/socket.h>
#include <linux/kernel.h>
#include <linux/string.h>
#include <linux/sockios.h>
#include <linux/net.h>
#include <linux/skbuff.h>
#include <asm/uaccess.h>
#include <net/sock.h>
#include <linux/init.h>

#include <net/netlink.h>
#include <linux/netfilter/nfnetlink.h>

MODULE_LICENSE("GPL");
MODULE_AUTHOR("Harald Welte <laforge@netfilter.org>");
MODULE_ALIAS_NET_PF_PROTO(PF_NETLINK, NETLINK_NETFILTER);

#define nfnl_dereference_protected(id) \
	rcu_dereference_protected(table[(id)].subsys, \
				  lockdep_nfnl_is_held((id)))

static char __initdata nfversion[] = "0.30";

static struct {
	struct mutex				mutex;
	const struct nfnetlink_subsystem __rcu	*subsys;
} table[NFNL_SUBSYS_COUNT];

static const int nfnl_group2type[NFNLGRP_MAX+1] = {
	[NFNLGRP_CONNTRACK_NEW]		= NFNL_SUBSYS_CTNETLINK,
	[NFNLGRP_CONNTRACK_UPDATE]	= NFNL_SUBSYS_CTNETLINK,
	[NFNLGRP_CONNTRACK_DESTROY]	= NFNL_SUBSYS_CTNETLINK,
	[NFNLGRP_CONNTRACK_EXP_NEW]	= NFNL_SUBSYS_CTNETLINK_EXP,
	[NFNLGRP_CONNTRACK_EXP_UPDATE]	= NFNL_SUBSYS_CTNETLINK_EXP,
	[NFNLGRP_CONNTRACK_EXP_DESTROY] = NFNL_SUBSYS_CTNETLINK_EXP,
	[NFNLGRP_NFTABLES]		= NFNL_SUBSYS_NFTABLES,
	[NFNLGRP_ACCT_QUOTA]		= NFNL_SUBSYS_ACCT,
	[NFNLGRP_NFTRACE]		= NFNL_SUBSYS_NFTABLES,
};

void nfnl_lock(__u8 subsys_id)
{
	mutex_lock(&table[subsys_id].mutex);
}
EXPORT_SYMBOL_GPL(nfnl_lock);

void nfnl_unlock(__u8 subsys_id)
{
	mutex_unlock(&table[subsys_id].mutex);
}
EXPORT_SYMBOL_GPL(nfnl_unlock);

#ifdef CONFIG_PROVE_LOCKING
bool lockdep_nfnl_is_held(u8 subsys_id)
{
	return lockdep_is_held(&table[subsys_id].mutex);
}
EXPORT_SYMBOL_GPL(lockdep_nfnl_is_held);
#endif

int nfnetlink_subsys_register(const struct nfnetlink_subsystem *n)
{
	nfnl_lock(n->subsys_id);
	if (table[n->subsys_id].subsys) {
		nfnl_unlock(n->subsys_id);
		return -EBUSY;
	}
	rcu_assign_pointer(table[n->subsys_id].subsys, n);
	nfnl_unlock(n->subsys_id);

	return 0;
}
EXPORT_SYMBOL_GPL(nfnetlink_subsys_register);

int nfnetlink_subsys_unregister(const struct nfnetlink_subsystem *n)
{
	nfnl_lock(n->subsys_id);
	table[n->subsys_id].subsys = NULL;
	nfnl_unlock(n->subsys_id);
	synchronize_rcu();
	return 0;
}
EXPORT_SYMBOL_GPL(nfnetlink_subsys_unregister);

static inline const struct nfnetlink_subsystem *nfnetlink_get_subsys(u_int16_t type)
{
	u_int8_t subsys_id = NFNL_SUBSYS_ID(type);

	if (subsys_id >= NFNL_SUBSYS_COUNT)
		return NULL;

	return rcu_dereference(table[subsys_id].subsys);
}

static inline const struct nfnl_callback *
nfnetlink_find_client(u_int16_t type, const struct nfnetlink_subsystem *ss)
{
	u_int8_t cb_id = NFNL_MSG_TYPE(type);

	if (cb_id >= ss->cb_count)
		return NULL;

	return &ss->cb[cb_id];
}

int nfnetlink_has_listeners(struct net *net, unsigned int group)
{
	return netlink_has_listeners(net->nfnl, group);
}
EXPORT_SYMBOL_GPL(nfnetlink_has_listeners);

int nfnetlink_send(struct sk_buff *skb, struct net *net, u32 portid,
		   unsigned int group, int echo, gfp_t flags)
{
	return nlmsg_notify(net->nfnl, skb, portid, group, echo, flags);
}
EXPORT_SYMBOL_GPL(nfnetlink_send);

int nfnetlink_set_err(struct net *net, u32 portid, u32 group, int error)
{
	return netlink_set_err(net->nfnl, portid, group, error);
}
EXPORT_SYMBOL_GPL(nfnetlink_set_err);

int nfnetlink_unicast(struct sk_buff *skb, struct net *net, u32 portid,
		      int flags)
{
	return netlink_unicast(net->nfnl, skb, portid, flags);
}
EXPORT_SYMBOL_GPL(nfnetlink_unicast);

/* Process one complete nfnetlink message. */
static int nfnetlink_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh)
{
	struct net *net = sock_net(skb->sk);
	const struct nfnl_callback *nc;
	const struct nfnetlink_subsystem *ss;
	int type, err;

	/* All the messages must at least contain nfgenmsg */
	if (nlmsg_len(nlh) < sizeof(struct nfgenmsg))
		return 0;

	type = nlh->nlmsg_type;
replay:
	rcu_read_lock();
	ss = nfnetlink_get_subsys(type);
	if (!ss) {
#ifdef CONFIG_MODULES
		rcu_read_unlock();
		request_module("nfnetlink-subsys-%d", NFNL_SUBSYS_ID(type));
		rcu_read_lock();
		ss = nfnetlink_get_subsys(type);
		if (!ss)
#endif
		{
			rcu_read_unlock();
			return -EINVAL;
		}
	}

	nc = nfnetlink_find_client(type, ss);
	if (!nc) {
		rcu_read_unlock();
		return -EINVAL;
	}

	{
		int min_len = nlmsg_total_size(sizeof(struct nfgenmsg));
		u_int8_t cb_id = NFNL_MSG_TYPE(nlh->nlmsg_type);
		struct nlattr *cda[ss->cb[cb_id].attr_count + 1];
		struct nlattr *attr = (void *)nlh + min_len;
		int attrlen = nlh->nlmsg_len - min_len;
		__u8 subsys_id = NFNL_SUBSYS_ID(type);

		err = nla_parse(cda, ss->cb[cb_id].attr_count,
				attr, attrlen, ss->cb[cb_id].policy);
		if (err < 0) {
			rcu_read_unlock();
			return err;
		}

		if (nc->call_rcu) {
			err = nc->call_rcu(net, net->nfnl, skb, nlh,
					   (const struct nlattr **)cda);
			rcu_read_unlock();
		} else {
			rcu_read_unlock();
			nfnl_lock(subsys_id);
			if (nfnl_dereference_protected(subsys_id) != ss ||
			    nfnetlink_find_client(type, ss) != nc)
				err = -EAGAIN;
			else if (nc->call)
				err = nc->call(net, net->nfnl, skb, nlh,
					       (const struct nlattr **)cda);
			else
				err = -EINVAL;
			nfnl_unlock(subsys_id);
		}
		if (err == -EAGAIN)
			goto replay;
		return err;
	}
}

struct nfnl_err {
	struct list_head	head;
	struct nlmsghdr		*nlh;
	int			err;
};

static int nfnl_err_add(struct list_head *list, struct nlmsghdr *nlh, int err)
{
	struct nfnl_err *nfnl_err;

	nfnl_err = kmalloc(sizeof(struct nfnl_err), GFP_KERNEL);
	if (nfnl_err == NULL)
		return -ENOMEM;

	nfnl_err->nlh = nlh;
	nfnl_err->err = err;
	list_add_tail(&nfnl_err->head, list);

	return 0;
}

static void nfnl_err_del(struct nfnl_err *nfnl_err)
{
	list_del(&nfnl_err->head);
	kfree(nfnl_err);
}

static void nfnl_err_reset(struct list_head *err_list)
{
	struct nfnl_err *nfnl_err, *next;

	list_for_each_entry_safe(nfnl_err, next, err_list, head)
		nfnl_err_del(nfnl_err);
}

static void nfnl_err_deliver(struct list_head *err_list, struct sk_buff *skb)
{
	struct nfnl_err *nfnl_err, *next;

	list_for_each_entry_safe(nfnl_err, next, err_list, head) {
		netlink_ack(skb, nfnl_err->nlh, nfnl_err->err);
		nfnl_err_del(nfnl_err);
	}
}

enum {
	NFNL_BATCH_FAILURE	= (1 << 0),
	NFNL_BATCH_DONE		= (1 << 1),
	NFNL_BATCH_REPLAY	= (1 << 2),
};

static void nfnetlink_rcv_batch(struct sk_buff *skb, struct nlmsghdr *nlh,
				u_int16_t subsys_id)
{
	struct sk_buff *oskb = skb;
	struct net *net = sock_net(skb->sk);
	const struct nfnetlink_subsystem *ss;
	const struct nfnl_callback *nc;
	static LIST_HEAD(err_list);
	u32 status;
	int err;

	if (subsys_id >= NFNL_SUBSYS_COUNT)
		return netlink_ack(skb, nlh, -EINVAL);
replay:
	status = 0;

	skb = netlink_skb_clone(oskb, GFP_KERNEL);
	if (!skb)
		return netlink_ack(oskb, nlh, -ENOMEM);

	nfnl_lock(subsys_id);
	ss = nfnl_dereference_protected(subsys_id);
	if (!ss) {
#ifdef CONFIG_MODULES
		nfnl_unlock(subsys_id);
		request_module("nfnetlink-subsys-%d", subsys_id);
		nfnl_lock(subsys_id);
		ss = nfnl_dereference_protected(subsys_id);
		if (!ss)
#endif
		{
			nfnl_unlock(subsys_id);
			netlink_ack(oskb, nlh, -EOPNOTSUPP);
			return kfree_skb(skb);
		}
	}

	if (!ss->commit || !ss->abort) {
		nfnl_unlock(subsys_id);
		netlink_ack(oskb, nlh, -EOPNOTSUPP);
		return kfree_skb(skb);
	}

	while (skb->len >= nlmsg_total_size(0)) {
		int msglen, type;

		nlh = nlmsg_hdr(skb);
		err = 0;

		if (nlh->nlmsg_len < NLMSG_HDRLEN ||
		    skb->len < nlh->nlmsg_len ||
		    nlmsg_len(nlh) < sizeof(struct nfgenmsg)) {
			nfnl_err_reset(&err_list);
			status |= NFNL_BATCH_FAILURE;
			goto done;
		}

		/* Only requests are handled by the kernel */
		if (!(nlh->nlmsg_flags & NLM_F_REQUEST)) {
			err = -EINVAL;
			goto ack;
		}

		type = nlh->nlmsg_type;
		if (type == NFNL_MSG_BATCH_BEGIN) {
			/* Malformed: Batch begin twice */
			nfnl_err_reset(&err_list);
			status |= NFNL_BATCH_FAILURE;
			goto done;
		} else if (type == NFNL_MSG_BATCH_END) {
			status |= NFNL_BATCH_DONE;
			goto done;
		} else if (type < NLMSG_MIN_TYPE) {
			err = -EINVAL;
			goto ack;
		}

		/* We only accept a batch with messages for the same
		 * subsystem.
		 */
		if (NFNL_SUBSYS_ID(type) != subsys_id) {
			err = -EINVAL;
			goto ack;
		}

		nc = nfnetlink_find_client(type, ss);
		if (!nc) {
			err = -EINVAL;
			goto ack;
		}

		{
			int min_len = nlmsg_total_size(sizeof(struct nfgenmsg));
			u_int8_t cb_id = NFNL_MSG_TYPE(nlh->nlmsg_type);
			struct nlattr *cda[ss->cb[cb_id].attr_count + 1];
			struct nlattr *attr = (void *)nlh + min_len;
			int attrlen = nlh->nlmsg_len - min_len;

			err = nla_parse(cda, ss->cb[cb_id].attr_count,
					attr, attrlen, ss->cb[cb_id].policy);
			if (err < 0)
				goto ack;

			if (nc->call_batch) {
				err = nc->call_batch(net, net->nfnl, skb, nlh,
						     (const struct nlattr **)cda);
			}

			/* The lock was released to autoload some module, we
			 * have to abort and start from scratch using the
			 * original skb.
			 */
			if (err == -EAGAIN) {
				status |= NFNL_BATCH_REPLAY;
				goto next;
			}
		}
ack:
		if (nlh->nlmsg_flags & NLM_F_ACK || err) {
			/* Errors are delivered once the full batch has been
			 * processed, this avoids that the same error is
			 * reported several times when replaying the batch.
			 */
			if (nfnl_err_add(&err_list, nlh, err) < 0) {
				/* We failed to enqueue an error, reset the
				 * list of errors and send OOM to userspace
				 * pointing to the batch header.
				 */
				nfnl_err_reset(&err_list);
				netlink_ack(oskb, nlmsg_hdr(oskb), -ENOMEM);
				status |= NFNL_BATCH_FAILURE;
				goto done;
			}
			/* We don't stop processing the batch on errors, thus,
			 * userspace gets all the errors that the batch
			 * triggers.
			 */
			if (err)
				status |= NFNL_BATCH_FAILURE;
		}
next:
		msglen = NLMSG_ALIGN(nlh->nlmsg_len);
		if (msglen > skb->len)
			msglen = skb->len;
		skb_pull(skb, msglen);
	}
done:
	if (status & NFNL_BATCH_REPLAY) {
		ss->abort(net, oskb);
		nfnl_err_reset(&err_list);
		nfnl_unlock(subsys_id);
		kfree_skb(skb);
		goto replay;
	} else if (status == NFNL_BATCH_DONE) {
		ss->commit(net, oskb);
	} else {
		ss->abort(net, oskb);
	}

	nfnl_err_deliver(&err_list, oskb);
	nfnl_unlock(subsys_id);
	kfree_skb(skb);
}

static void nfnetlink_rcv(struct sk_buff *skb)
{
	struct nlmsghdr *nlh = nlmsg_hdr(skb);
	u_int16_t res_id;
	int msglen;

	if (nlh->nlmsg_len < NLMSG_HDRLEN ||
	    skb->len < nlh->nlmsg_len)
		return;

	if (!netlink_net_capable(skb, CAP_NET_ADMIN)) {
		netlink_ack(skb, nlh, -EPERM);
		return;
	}

	if (nlh->nlmsg_type == NFNL_MSG_BATCH_BEGIN) {
		struct nfgenmsg *nfgenmsg;

		msglen = NLMSG_ALIGN(nlh->nlmsg_len);
		if (msglen > skb->len)
			msglen = skb->len;

		if (nlh->nlmsg_len < NLMSG_HDRLEN ||
		    skb->len < NLMSG_HDRLEN + sizeof(struct nfgenmsg))
			return;

		nfgenmsg = nlmsg_data(nlh);
		skb_pull(skb, msglen);
		/* Work around old nft using host byte order */
		if (nfgenmsg->res_id == NFNL_SUBSYS_NFTABLES)
			res_id = NFNL_SUBSYS_NFTABLES;
		else
			res_id = ntohs(nfgenmsg->res_id);
		nfnetlink_rcv_batch(skb, nlh, res_id);
	} else {
		netlink_rcv_skb(skb, &nfnetlink_rcv_msg);
	}
}

#ifdef CONFIG_MODULES
static int nfnetlink_bind(struct net *net, int group)
{
	const struct nfnetlink_subsystem *ss;
	int type;

	if (group <= NFNLGRP_NONE || group > NFNLGRP_MAX)
		return 0;

	type = nfnl_group2type[group];

	rcu_read_lock();
	ss = nfnetlink_get_subsys(type << 8);
	rcu_read_unlock();
	if (!ss)
		request_module("nfnetlink-subsys-%d", type);
	return 0;
}
#endif

static int __net_init nfnetlink_net_init(struct net *net)
{
	struct sock *nfnl;
	struct netlink_kernel_cfg cfg = {
		.groups	= NFNLGRP_MAX,
		.input	= nfnetlink_rcv,
#ifdef CONFIG_MODULES
		.bind	= nfnetlink_bind,
#endif
	};

	nfnl = netlink_kernel_create(net, NETLINK_NETFILTER, &cfg);
	if (!nfnl)
		return -ENOMEM;
	net->nfnl_stash = nfnl;
	rcu_assign_pointer(net->nfnl, nfnl);
	return 0;
}

static void __net_exit nfnetlink_net_exit_batch(struct list_head *net_exit_list)
{
	struct net *net;

	list_for_each_entry(net, net_exit_list, exit_list)
		RCU_INIT_POINTER(net->nfnl, NULL);
	synchronize_net();
	list_for_each_entry(net, net_exit_list, exit_list)
		netlink_kernel_release(net->nfnl_stash);
}

static struct pernet_operations nfnetlink_net_ops = {
	.init		= nfnetlink_net_init,
	.exit_batch	= nfnetlink_net_exit_batch,
};

static int __init nfnetlink_init(void)
{
	int i;

	for (i = NFNLGRP_NONE + 1; i <= NFNLGRP_MAX; i++)
		BUG_ON(nfnl_group2type[i] == NFNL_SUBSYS_NONE);

	for (i=0; i<NFNL_SUBSYS_COUNT; i++)
		mutex_init(&table[i].mutex);

	pr_info("Netfilter messages via NETLINK v%s.\n", nfversion);
	return register_pernet_subsys(&nfnetlink_net_ops);
}

static void __exit nfnetlink_exit(void)
{
	pr_info("Removing netfilter NETLINK layer.\n");
	unregister_pernet_subsys(&nfnetlink_net_ops);
}
module_init(nfnetlink_init);
module_exit(nfnetlink_exit);
Commit	Line	Data
f9e815b3 HW	1	/* Netfilter messages via netlink socket. Allows for user space
	2	* protocol helpers and general trouble making from userspace.
	3	*
	4	* (C) 2001 by Jay Schulist <jschlst@samba.org>,
	5	* (C) 2002-2005 by Harald Welte <laforge@gnumonks.org>
67ca3966	6	* (C) 2005,2007 by Pablo Neira Ayuso <pablo@netfilter.org>
f9e815b3 HW	7	*
	8	* Initial netfilter messages via netlink development funded and
	9	* generally made possible by Network Robots, Inc. (www.networkrobots.com)
	10	*
	11	* Further development of this code funded by Astaro AG (http://www.astaro.com)
	12	*
	13	* This software may be used and distributed according to the terms
	14	* of the GNU General Public License, incorporated herein by reference.
	15	*/
	16
f9e815b3 HW	17	#include <linux/module.h>
	18	#include <linux/types.h>
	19	#include <linux/socket.h>
	20	#include <linux/kernel.h>
f9e815b3 HW	21	#include <linux/string.h>
	22	#include <linux/sockios.h>
	23	#include <linux/net.h>
f9e815b3 HW	24	#include <linux/skbuff.h>
f9e815b3 HW	25	#include <asm/uaccess.h>
f9e815b3 HW	26	#include <net/sock.h>
f9e815b3 HW	27	#include <linux/init.h>
f9e815b3	28
573ce260	29	#include <net/netlink.h>
f9e815b3 HW	30	#include <linux/netfilter/nfnetlink.h>
	31
	32	MODULE_LICENSE("GPL");
4fdb3bb7 HW	33	MODULE_AUTHOR("Harald Welte <laforge@netfilter.org>");
4fdb3bb7 HW	34	MODULE_ALIAS_NET_PF_PROTO(PF_NETLINK, NETLINK_NETFILTER);
f9e815b3	35
9c55d3b5 FW	36	#define nfnl_dereference_protected(id) \
	37	rcu_dereference_protected(table[(id)].subsys, \
	38	lockdep_nfnl_is_held((id)))
	39
f9e815b3 HW	40	static char __initdata nfversion[] = "0.30";
f9e815b3 HW	41
c14b78e7 PNA	42	static struct {
	43	struct mutex mutex;
	44	const struct nfnetlink_subsystem __rcu *subsys;
	45	} table[NFNL_SUBSYS_COUNT];
f9e815b3	46
03292745 PNA	47	static const int nfnl_group2type[NFNLGRP_MAX+1] = {
	48	[NFNLGRP_CONNTRACK_NEW] = NFNL_SUBSYS_CTNETLINK,
	49	[NFNLGRP_CONNTRACK_UPDATE] = NFNL_SUBSYS_CTNETLINK,
	50	[NFNLGRP_CONNTRACK_DESTROY] = NFNL_SUBSYS_CTNETLINK,
	51	[NFNLGRP_CONNTRACK_EXP_NEW] = NFNL_SUBSYS_CTNETLINK_EXP,
	52	[NFNLGRP_CONNTRACK_EXP_UPDATE] = NFNL_SUBSYS_CTNETLINK_EXP,
	53	[NFNLGRP_CONNTRACK_EXP_DESTROY] = NFNL_SUBSYS_CTNETLINK_EXP,
97840cb6 PNA	54	[NFNLGRP_NFTABLES] = NFNL_SUBSYS_NFTABLES,
97840cb6 PNA	55	[NFNLGRP_ACCT_QUOTA] = NFNL_SUBSYS_ACCT,
33d5a7b1	56	[NFNLGRP_NFTRACE] = NFNL_SUBSYS_NFTABLES,
03292745 PNA	57	};
03292745 PNA	58
c14b78e7	59	void nfnl_lock(__u8 subsys_id)
f9e815b3	60	{
c14b78e7	61	mutex_lock(&table[subsys_id].mutex);
f9e815b3	62	}
e6a7d3c0	63	EXPORT_SYMBOL_GPL(nfnl_lock);
f9e815b3	64
c14b78e7	65	void nfnl_unlock(__u8 subsys_id)
a3c5029c	66	{
c14b78e7	67	mutex_unlock(&table[subsys_id].mutex);
f9e815b3	68	}
e6a7d3c0	69	EXPORT_SYMBOL_GPL(nfnl_unlock);
f9e815b3	70
0eb5db7a	71	#ifdef CONFIG_PROVE_LOCKING
875e0829	72	bool lockdep_nfnl_is_held(u8 subsys_id)
0eb5db7a PM	73	{
	74	return lockdep_is_held(&table[subsys_id].mutex);
	75	}
	76	EXPORT_SYMBOL_GPL(lockdep_nfnl_is_held);
	77	#endif
	78
7c8d4cb4	79	int nfnetlink_subsys_register(const struct nfnetlink_subsystem *n)
f9e815b3	80	{
c14b78e7 PNA	81	nfnl_lock(n->subsys_id);
	82	if (table[n->subsys_id].subsys) {
	83	nfnl_unlock(n->subsys_id);
0ab43f84 HW	84	return -EBUSY;
0ab43f84 HW	85	}
c14b78e7 PNA	86	rcu_assign_pointer(table[n->subsys_id].subsys, n);
c14b78e7 PNA	87	nfnl_unlock(n->subsys_id);
f9e815b3 HW	88
	89	return 0;
	90	}
f4bc177f	91	EXPORT_SYMBOL_GPL(nfnetlink_subsys_register);
f9e815b3	92
7c8d4cb4	93	int nfnetlink_subsys_unregister(const struct nfnetlink_subsystem *n)
f9e815b3	94	{
c14b78e7 PNA	95	nfnl_lock(n->subsys_id);
	96	table[n->subsys_id].subsys = NULL;
	97	nfnl_unlock(n->subsys_id);
6b75e3e8	98	synchronize_rcu();
f9e815b3 HW	99	return 0;
f9e815b3 HW	100	}
f4bc177f	101	EXPORT_SYMBOL_GPL(nfnetlink_subsys_unregister);
f9e815b3	102
7c8d4cb4	103	static inline const struct nfnetlink_subsystem *nfnetlink_get_subsys(u_int16_t type)
f9e815b3 HW	104	{
	105	u_int8_t subsys_id = NFNL_SUBSYS_ID(type);
	106
ac0f1d98	107	if (subsys_id >= NFNL_SUBSYS_COUNT)
f9e815b3 HW	108	return NULL;
f9e815b3 HW	109
c14b78e7	110	return rcu_dereference(table[subsys_id].subsys);
f9e815b3 HW	111	}
f9e815b3 HW	112
7c8d4cb4 PM	113	static inline const struct nfnl_callback *
7c8d4cb4 PM	114	nfnetlink_find_client(u_int16_t type, const struct nfnetlink_subsystem *ss)
f9e815b3 HW	115	{
f9e815b3 HW	116	u_int8_t cb_id = NFNL_MSG_TYPE(type);
601e68e1	117
67ca3966	118	if (cb_id >= ss->cb_count)
f9e815b3	119	return NULL;
f9e815b3 HW	120
	121	return &ss->cb[cb_id];
	122	}
	123
cd8c20b6	124	int nfnetlink_has_listeners(struct net *net, unsigned int group)
a2427692	125	{
cd8c20b6	126	return netlink_has_listeners(net->nfnl, group);
a2427692 PM	127	}
	128	EXPORT_SYMBOL_GPL(nfnetlink_has_listeners);
	129
ec464e5d	130	int nfnetlink_send(struct sk_buff skb, struct net net, u32 portid,
95c96174	131	unsigned int group, int echo, gfp_t flags)
f9e815b3	132	{
ec464e5d	133	return nlmsg_notify(net->nfnl, skb, portid, group, echo, flags);
f9e815b3	134	}
f4bc177f	135	EXPORT_SYMBOL_GPL(nfnetlink_send);
f9e815b3	136
ec464e5d	137	int nfnetlink_set_err(struct net *net, u32 portid, u32 group, int error)
dd5b6ce6	138	{
ec464e5d	139	return netlink_set_err(net->nfnl, portid, group, error);
dd5b6ce6 PNA	140	}
	141	EXPORT_SYMBOL_GPL(nfnetlink_set_err);
	142
ec464e5d PM	143	int nfnetlink_unicast(struct sk_buff skb, struct net net, u32 portid,
ec464e5d PM	144	int flags)
f9e815b3	145	{
ec464e5d	146	return netlink_unicast(net->nfnl, skb, portid, flags);
f9e815b3	147	}
f4bc177f	148	EXPORT_SYMBOL_GPL(nfnetlink_unicast);
f9e815b3 HW	149
f9e815b3 HW	150	/* Process one complete nfnetlink message. */
1d00a4eb	151	static int nfnetlink_rcv_msg(struct sk_buff skb, struct nlmsghdr nlh)
f9e815b3	152	{
cd8c20b6	153	struct net *net = sock_net(skb->sk);
7c8d4cb4 PM	154	const struct nfnl_callback *nc;
7c8d4cb4 PM	155	const struct nfnetlink_subsystem *ss;
1d00a4eb	156	int type, err;
f9e815b3	157
f9e815b3	158	/* All the messages must at least contain nfgenmsg */
573ce260	159	if (nlmsg_len(nlh) < sizeof(struct nfgenmsg))
f9e815b3	160	return 0;
f9e815b3 HW	161
f9e815b3 HW	162	type = nlh->nlmsg_type;
e6a7d3c0	163	replay:
6b75e3e8	164	rcu_read_lock();
f9e815b3	165	ss = nfnetlink_get_subsys(type);
0ab43f84	166	if (!ss) {
95a5afca	167	#ifdef CONFIG_MODULES
6b75e3e8	168	rcu_read_unlock();
37d2e7a2	169	request_module("nfnetlink-subsys-%d", NFNL_SUBSYS_ID(type));
6b75e3e8	170	rcu_read_lock();
37d2e7a2	171	ss = nfnetlink_get_subsys(type);
0ab43f84 HW	172	if (!ss)
0ab43f84 HW	173	#endif
6b75e3e8 ED	174	{
6b75e3e8 ED	175	rcu_read_unlock();
1d00a4eb	176	return -EINVAL;
6b75e3e8	177	}
0ab43f84	178	}
f9e815b3 HW	179
f9e815b3 HW	180	nc = nfnetlink_find_client(type, ss);
6b75e3e8 ED	181	if (!nc) {
6b75e3e8 ED	182	rcu_read_unlock();
1d00a4eb	183	return -EINVAL;
6b75e3e8	184	}
f9e815b3	185
f9e815b3	186	{
573ce260	187	int min_len = nlmsg_total_size(sizeof(struct nfgenmsg));
e3730578	188	u_int8_t cb_id = NFNL_MSG_TYPE(nlh->nlmsg_type);
f49c857f PNA	189	struct nlattr *cda[ss->cb[cb_id].attr_count + 1];
	190	struct nlattr attr = (void )nlh + min_len;
	191	int attrlen = nlh->nlmsg_len - min_len;
c14b78e7	192	__u8 subsys_id = NFNL_SUBSYS_ID(type);
f49c857f PNA	193
	194	err = nla_parse(cda, ss->cb[cb_id].attr_count,
	195	attr, attrlen, ss->cb[cb_id].policy);
4009e188 TB	196	if (err < 0) {
4009e188 TB	197	rcu_read_unlock();
f49c857f	198	return err;
4009e188	199	}
601e68e1	200
6b75e3e8	201	if (nc->call_rcu) {
7b8002a1	202	err = nc->call_rcu(net, net->nfnl, skb, nlh,
6b75e3e8 ED	203	(const struct nlattr **)cda);
	204	rcu_read_unlock();
	205	} else {
	206	rcu_read_unlock();
c14b78e7	207	nfnl_lock(subsys_id);
9c55d3b5	208	if (nfnl_dereference_protected(subsys_id) != ss \|\|
6b75e3e8 ED	209	nfnetlink_find_client(type, ss) != nc)
6b75e3e8 ED	210	err = -EAGAIN;
59560a38	211	else if (nc->call)
7b8002a1 PNA	212	err = nc->call(net, net->nfnl, skb, nlh,
7b8002a1 PNA	213	(const struct nlattr **)cda);
59560a38 TB	214	else
59560a38 TB	215	err = -EINVAL;
c14b78e7	216	nfnl_unlock(subsys_id);
6b75e3e8	217	}
e6a7d3c0 PNA	218	if (err == -EAGAIN)
	219	goto replay;
	220	return err;
f9e815b3	221	}
f9e815b3 HW	222	}
f9e815b3 HW	223
cbb8125e PNA	224	struct nfnl_err {
	225	struct list_head head;
	226	struct nlmsghdr *nlh;
	227	int err;
	228	};
	229
	230	static int nfnl_err_add(struct list_head list, struct nlmsghdr nlh, int err)
	231	{
	232	struct nfnl_err *nfnl_err;
	233
	234	nfnl_err = kmalloc(sizeof(struct nfnl_err), GFP_KERNEL);
	235	if (nfnl_err == NULL)
	236	return -ENOMEM;
	237
	238	nfnl_err->nlh = nlh;
	239	nfnl_err->err = err;
	240	list_add_tail(&nfnl_err->head, list);
	241
	242	return 0;
	243	}
	244
	245	static void nfnl_err_del(struct nfnl_err *nfnl_err)
	246	{
	247	list_del(&nfnl_err->head);
	248	kfree(nfnl_err);
	249	}
	250
	251	static void nfnl_err_reset(struct list_head *err_list)
	252	{
	253	struct nfnl_err nfnl_err, next;
	254
	255	list_for_each_entry_safe(nfnl_err, next, err_list, head)
	256	nfnl_err_del(nfnl_err);
	257	}
	258
	259	static void nfnl_err_deliver(struct list_head err_list, struct sk_buff skb)
	260	{
	261	struct nfnl_err nfnl_err, next;
	262
	263	list_for_each_entry_safe(nfnl_err, next, err_list, head) {
	264	netlink_ack(skb, nfnl_err->nlh, nfnl_err->err);
	265	nfnl_err_del(nfnl_err);
	266	}
	267	}
	268
6742b9e3 PNA	269	enum {
	270	NFNL_BATCH_FAILURE = (1 << 0),
	271	NFNL_BATCH_DONE = (1 << 1),
	272	NFNL_BATCH_REPLAY = (1 << 2),
	273	};
	274
0628b123 PNA	275	static void nfnetlink_rcv_batch(struct sk_buff skb, struct nlmsghdr nlh,
	276	u_int16_t subsys_id)
	277	{
0f816232	278	struct sk_buff *oskb = skb;
0628b123 PNA	279	struct net *net = sock_net(skb->sk);
	280	const struct nfnetlink_subsystem *ss;
	281	const struct nfnl_callback *nc;
cbb8125e	282	static LIST_HEAD(err_list);
6742b9e3	283	u32 status;
0628b123 PNA	284	int err;
	285
	286	if (subsys_id >= NFNL_SUBSYS_COUNT)
	287	return netlink_ack(skb, nlh, -EINVAL);
	288	replay:
6742b9e3 PNA	289	status = 0;
6742b9e3 PNA	290
0f816232 DJ	291	skb = netlink_skb_clone(oskb, GFP_KERNEL);
0f816232 DJ	292	if (!skb)
0628b123 PNA	293	return netlink_ack(oskb, nlh, -ENOMEM);
0628b123 PNA	294
0628b123	295	nfnl_lock(subsys_id);
9c55d3b5	296	ss = nfnl_dereference_protected(subsys_id);
0628b123 PNA	297	if (!ss) {
	298	#ifdef CONFIG_MODULES
	299	nfnl_unlock(subsys_id);
	300	request_module("nfnetlink-subsys-%d", subsys_id);
	301	nfnl_lock(subsys_id);
9c55d3b5	302	ss = nfnl_dereference_protected(subsys_id);
0628b123 PNA	303	if (!ss)
	304	#endif
	305	{
	306	nfnl_unlock(subsys_id);
7c7bdf35	307	netlink_ack(oskb, nlh, -EOPNOTSUPP);
0f816232	308	return kfree_skb(skb);
0628b123 PNA	309	}
	310	}
	311
	312	if (!ss->commit \|\| !ss->abort) {
	313	nfnl_unlock(subsys_id);
7c7bdf35	314	netlink_ack(oskb, nlh, -EOPNOTSUPP);
ecd15dd7	315	return kfree_skb(skb);
0628b123 PNA	316	}
	317
	318	while (skb->len >= nlmsg_total_size(0)) {
	319	int msglen, type;
	320
	321	nlh = nlmsg_hdr(skb);
	322	err = 0;
	323
c58d6c93 PT	324	if (nlh->nlmsg_len < NLMSG_HDRLEN \|\|
	325	skb->len < nlh->nlmsg_len \|\|
	326	nlmsg_len(nlh) < sizeof(struct nfgenmsg)) {
	327	nfnl_err_reset(&err_list);
	328	status \|= NFNL_BATCH_FAILURE;
	329	goto done;
0628b123 PNA	330	}
	331
	332	/* Only requests are handled by the kernel */
	333	if (!(nlh->nlmsg_flags & NLM_F_REQUEST)) {
	334	err = -EINVAL;
	335	goto ack;
	336	}
	337
	338	type = nlh->nlmsg_type;
	339	if (type == NFNL_MSG_BATCH_BEGIN) {
	340	/* Malformed: Batch begin twice */
cbb8125e	341	nfnl_err_reset(&err_list);
6742b9e3	342	status \|= NFNL_BATCH_FAILURE;
0628b123 PNA	343	goto done;
0628b123 PNA	344	} else if (type == NFNL_MSG_BATCH_END) {
6742b9e3	345	status \|= NFNL_BATCH_DONE;
0628b123 PNA	346	goto done;
	347	} else if (type < NLMSG_MIN_TYPE) {
	348	err = -EINVAL;
	349	goto ack;
	350	}
	351
	352	/* We only accept a batch with messages for the same
	353	* subsystem.
	354	*/
	355	if (NFNL_SUBSYS_ID(type) != subsys_id) {
	356	err = -EINVAL;
	357	goto ack;
	358	}
	359
	360	nc = nfnetlink_find_client(type, ss);
	361	if (!nc) {
	362	err = -EINVAL;
	363	goto ack;
	364	}
	365
	366	{
	367	int min_len = nlmsg_total_size(sizeof(struct nfgenmsg));
	368	u_int8_t cb_id = NFNL_MSG_TYPE(nlh->nlmsg_type);
	369	struct nlattr *cda[ss->cb[cb_id].attr_count + 1];
	370	struct nlattr attr = (void )nlh + min_len;
	371	int attrlen = nlh->nlmsg_len - min_len;
	372
	373	err = nla_parse(cda, ss->cb[cb_id].attr_count,
	374	attr, attrlen, ss->cb[cb_id].policy);
	375	if (err < 0)
	376	goto ack;
	377
	378	if (nc->call_batch) {
633c9a84	379	err = nc->call_batch(net, net->nfnl, skb, nlh,
0628b123 PNA	380	(const struct nlattr **)cda);
	381	}
	382
	383	/* The lock was released to autoload some module, we
	384	* have to abort and start from scratch using the
	385	* original skb.
	386	*/
	387	if (err == -EAGAIN) {
6742b9e3 PNA	388	status \|= NFNL_BATCH_REPLAY;
6742b9e3 PNA	389	goto next;
0628b123 PNA	390	}
	391	}
	392	ack:
	393	if (nlh->nlmsg_flags & NLM_F_ACK \|\| err) {
cbb8125e PNA	394	/* Errors are delivered once the full batch has been
	395	* processed, this avoids that the same error is
	396	* reported several times when replaying the batch.
	397	*/
	398	if (nfnl_err_add(&err_list, nlh, err) < 0) {
	399	/* We failed to enqueue an error, reset the
	400	* list of errors and send OOM to userspace
	401	* pointing to the batch header.
	402	*/
	403	nfnl_err_reset(&err_list);
7c7bdf35	404	netlink_ack(oskb, nlmsg_hdr(oskb), -ENOMEM);
6742b9e3	405	status \|= NFNL_BATCH_FAILURE;
cbb8125e PNA	406	goto done;
cbb8125e PNA	407	}
0628b123 PNA	408	/* We don't stop processing the batch on errors, thus,
	409	* userspace gets all the errors that the batch
	410	* triggers.
	411	*/
0628b123	412	if (err)
6742b9e3	413	status \|= NFNL_BATCH_FAILURE;
0628b123	414	}
6742b9e3	415	next:
0628b123 PNA	416	msglen = NLMSG_ALIGN(nlh->nlmsg_len);
	417	if (msglen > skb->len)
	418	msglen = skb->len;
	419	skb_pull(skb, msglen);
	420	}
	421	done:
6742b9e3	422	if (status & NFNL_BATCH_REPLAY) {
5913beaf	423	ss->abort(net, oskb);
6742b9e3 PNA	424	nfnl_err_reset(&err_list);
	425	nfnl_unlock(subsys_id);
	426	kfree_skb(skb);
	427	goto replay;
	428	} else if (status == NFNL_BATCH_DONE) {
5913beaf	429	ss->commit(net, oskb);
6742b9e3	430	} else {
5913beaf	431	ss->abort(net, oskb);
6742b9e3	432	}
0628b123	433
cbb8125e	434	nfnl_err_deliver(&err_list, oskb);
0628b123	435	nfnl_unlock(subsys_id);
0f816232	436	kfree_skb(skb);
0628b123 PNA	437	}
0628b123 PNA	438
cd40b7d3	439	static void nfnetlink_rcv(struct sk_buff *skb)
f9e815b3	440	{
0628b123	441	struct nlmsghdr *nlh = nlmsg_hdr(skb);
a9de9777	442	u_int16_t res_id;
0628b123 PNA	443	int msglen;
0628b123 PNA	444
0628b123 PNA	445	if (nlh->nlmsg_len < NLMSG_HDRLEN \|\|
	446	skb->len < nlh->nlmsg_len)
	447	return;
	448
90f62cf3	449	if (!netlink_net_capable(skb, CAP_NET_ADMIN)) {
cdbe7c2d JB	450	netlink_ack(skb, nlh, -EPERM);
	451	return;
	452	}
	453
0628b123 PNA	454	if (nlh->nlmsg_type == NFNL_MSG_BATCH_BEGIN) {
	455	struct nfgenmsg *nfgenmsg;
	456
	457	msglen = NLMSG_ALIGN(nlh->nlmsg_len);
	458	if (msglen > skb->len)
	459	msglen = skb->len;
	460
	461	if (nlh->nlmsg_len < NLMSG_HDRLEN \|\|
	462	skb->len < NLMSG_HDRLEN + sizeof(struct nfgenmsg))
	463	return;
	464
	465	nfgenmsg = nlmsg_data(nlh);
	466	skb_pull(skb, msglen);
a9de9777 PNA	467	/* Work around old nft using host byte order */
	468	if (nfgenmsg->res_id == NFNL_SUBSYS_NFTABLES)
	469	res_id = NFNL_SUBSYS_NFTABLES;
	470	else
	471	res_id = ntohs(nfgenmsg->res_id);
	472	nfnetlink_rcv_batch(skb, nlh, res_id);
0628b123 PNA	473	} else {
	474	netlink_rcv_skb(skb, &nfnetlink_rcv_msg);
	475	}
f9e815b3 HW	476	}
f9e815b3 HW	477
03292745	478	#ifdef CONFIG_MODULES
023e2cfa	479	static int nfnetlink_bind(struct net *net, int group)
03292745 PNA	480	{
03292745 PNA	481	const struct nfnetlink_subsystem *ss;
97840cb6 PNA	482	int type;
	483
	484	if (group <= NFNLGRP_NONE \|\| group > NFNLGRP_MAX)
62924af2	485	return 0;
97840cb6 PNA	486
97840cb6 PNA	487	type = nfnl_group2type[group];
03292745 PNA	488
03292745 PNA	489	rcu_read_lock();
dbc3617f	490	ss = nfnetlink_get_subsys(type << 8);
03292745	491	rcu_read_unlock();
bfe4bc71 RGB	492	if (!ss)
bfe4bc71 RGB	493	request_module("nfnetlink-subsys-%d", type);
4f520900	494	return 0;
03292745 PNA	495	}
	496	#endif
	497
cd8c20b6	498	static int __net_init nfnetlink_net_init(struct net *net)
f9e815b3	499	{
cd8c20b6	500	struct sock *nfnl;
a31f2d17 PNA	501	struct netlink_kernel_cfg cfg = {
	502	.groups = NFNLGRP_MAX,
	503	.input = nfnetlink_rcv,
03292745 PNA	504	#ifdef CONFIG_MODULES
	505	.bind = nfnetlink_bind,
	506	#endif
a31f2d17	507	};
cd8c20b6	508
9f00d977	509	nfnl = netlink_kernel_create(net, NETLINK_NETFILTER, &cfg);
cd8c20b6 AD	510	if (!nfnl)
	511	return -ENOMEM;
	512	net->nfnl_stash = nfnl;
cf778b00	513	rcu_assign_pointer(net->nfnl, nfnl);
cd8c20b6	514	return 0;
f9e815b3 HW	515	}
f9e815b3 HW	516
cd8c20b6	517	static void __net_exit nfnetlink_net_exit_batch(struct list_head *net_exit_list)
f9e815b3	518	{
cd8c20b6	519	struct net *net;
f9e815b3	520
cd8c20b6	521	list_for_each_entry(net, net_exit_list, exit_list)
a9b3cd7f	522	RCU_INIT_POINTER(net->nfnl, NULL);
cd8c20b6 AD	523	synchronize_net();
	524	list_for_each_entry(net, net_exit_list, exit_list)
	525	netlink_kernel_release(net->nfnl_stash);
	526	}
f9e815b3	527
cd8c20b6 AD	528	static struct pernet_operations nfnetlink_net_ops = {
	529	.init = nfnetlink_net_init,
	530	.exit_batch = nfnetlink_net_exit_batch,
	531	};
	532
	533	static int __init nfnetlink_init(void)
	534	{
c14b78e7 PNA	535	int i;
c14b78e7 PNA	536
97840cb6 PNA	537	for (i = NFNLGRP_NONE + 1; i <= NFNLGRP_MAX; i++)
	538	BUG_ON(nfnl_group2type[i] == NFNL_SUBSYS_NONE);
	539
c14b78e7 PNA	540	for (i=0; i<NFNL_SUBSYS_COUNT; i++)
	541	mutex_init(&table[i].mutex);
	542
654d0fbd	543	pr_info("Netfilter messages via NETLINK v%s.\n", nfversion);
cd8c20b6	544	return register_pernet_subsys(&nfnetlink_net_ops);
f9e815b3 HW	545	}
f9e815b3 HW	546
cd8c20b6 AD	547	static void __exit nfnetlink_exit(void)
cd8c20b6 AD	548	{
654d0fbd	549	pr_info("Removing netfilter NETLINK layer.\n");
cd8c20b6 AD	550	unregister_pernet_subsys(&nfnetlink_net_ops);
cd8c20b6 AD	551	}
f9e815b3 HW	552	module_init(nfnetlink_init);
f9e815b3 HW	553	module_exit(nfnetlink_exit);